Perplexity LLM

Overview

PerplexityAgentHttpLLMService provides access to Perplexity’s Agent API, which answers with web search using models from Perplexity and other providers. It inherits from OpenAIResponsesHttpLLMService, so streaming, function calling, and context management work as they do on the OpenAI Responses page.

Perplexity LLM API Reference

Pipecat’s API methods for Perplexity integration

Example Implementation

Complete example with search capabilities

Perplexity Documentation

Official Perplexity API documentation and features

Perplexity Platform

Access search-enhanced models and API keys

Installation

To use Perplexity services, install the required dependencies:

uv add "pipecat-ai[perplexity]"

Prerequisites

Perplexity Account Setup

Before using Perplexity LLM services, you need:

Perplexity Account: Sign up at Perplexity
API Key: Generate an API key from your account dashboard
Model or Preset Selection: Choose a model, a preset, or both

Required Environment Variables

PERPLEXITY_API_KEY: Your Perplexity API key for authentication

Configuration

from pipecat.services.perplexity.llm import PerplexityAgentHttpLLMService

str

default:"None"

Perplexity API key for authentication.

str

default:"https://api.perplexity.ai/v1"

Base URL for Perplexity API endpoint.

PerplexityAgentHttpLLMService.Settings

default:"None"

Runtime-configurable settings. See Settings below.

Other constructor arguments, such as retry_on_timeout, are those of OpenAIResponsesHttpLLMService.

Settings

Runtime-configurable settings passed via the settings constructor argument using PerplexityAgentHttpLLMService.Settings(...). These can be updated mid-conversation with LLMUpdateSettingsFrame. See Service Settings for details. This service adds the following to the settings of OpenAIResponsesHttpLLMService (see OpenAI Responses Settings). The default model is "openai/gpt-5.6-luna".

Parameter	Type	Default	Description
`preset`	`str \| None`	`None`	Perplexity preset that sets the model, its instructions, its search budget, and how much it reasons: `"fast"`, `"low"`, `"medium"`, `"high"`, or `"xhigh"`. `model` replaces the preset’s model; set `model=None` to use the preset’s. At least one of `model` and `preset` must be set.
`web_search`	`bool`	`True`	Whether to offer the `web_search` tool on every request, alongside any tools from the context. A preset searches the web whatever this is set to.
`max_steps`	`int \| None`	`None`	Maximum number of search and reasoning steps per response. `None` uses Perplexity’s default.

Raises ValueError when the constructor settings leave neither model nor preset. A runtime update that leaves neither pushes an ErrorFrame. With neither preset nor reasoning set, each request sends the "fast" preset, Perplexity’s lowest-latency setup, and the first request for each model logs it. With use_provider_thinking_default set to True, no preset is sent.

Usage

Basic Setup

import os
from pipecat.services.perplexity.llm import PerplexityAgentHttpLLMService

llm = PerplexityAgentHttpLLMService(
    api_key=os.getenv("PERPLEXITY_API_KEY"),
    settings=PerplexityAgentHttpLLMService.Settings(
        system_instruction="You are a helpful assistant.",
    ),
)

With a Preset

llm = PerplexityAgentHttpLLMService(
    api_key=os.getenv("PERPLEXITY_API_KEY"),
    settings=PerplexityAgentHttpLLMService.Settings(
        model=None,
        preset="medium",
        max_steps=3,
    ),
)

Notes

Citations: Answers can include citation markers such as [1].
Anthropic models: Models from Anthropic need max_completion_tokens set in the settings.

Deprecated: PerplexityLLMService

Deprecated since 1.13.0, removed in 2.0.0. Use PerplexityAgentHttpLLMService instead. Perplexity has ended support for the Sonar chat completions API that PerplexityLLMService uses.

Get Started

Pipecat Server

Client SDKs

Pipecat Flows

Pipecat Cloud

CLI

Pipecat Context Hub

Overview

Perplexity LLM API Reference

Example Implementation

Perplexity Documentation

Perplexity Platform

Installation

Prerequisites

Perplexity Account Setup

Required Environment Variables

Configuration

Settings

Usage

Basic Setup

With a Preset

Notes

Deprecated: PerplexityLLMService

​Overview

Perplexity LLM API Reference

Example Implementation

Perplexity Documentation

Perplexity Platform

​Installation

​Prerequisites

​Perplexity Account Setup

​Required Environment Variables

​Configuration

​Settings

​Usage

​Basic Setup

​With a Preset

​Notes

​Deprecated: PerplexityLLMService

Overview

Installation

Prerequisites

Perplexity Account Setup

Required Environment Variables

Configuration

Settings

Usage

Basic Setup

With a Preset

Notes

Deprecated: PerplexityLLMService